调用m = TextToSpeech(speaker_model_path='shibing624/parrots-gpt-sovits-speaker-maimai', speaker_name='MaiMai') 报错

[bombhero]

第一步调用就报错了,我的pytorch版本是2.2.1+cu121, 是不是太高了?

Cell In[3], line 1
----> 1 m = TextToSpeech(speaker_model_path='shibing624/parrots-gpt-sovits-speaker-maimai', speaker_name='MaiMai')

File e:\bomb\proj\python\BarkVoice\parrots\tts.py:342, in TextToSpeech.init(self, bert_model_path, hubert_model_path, sovits_model_path, gpt_model_path, speaker_model_path, speaker_name, device, half)
339 raise ValueError("sovits_model_path, gpt_model_path or speaker_model_path must be provided")
341 # SoVITS
--> 342 sovits_dict = torch.load(sovits_model_path, map_location="cpu")
343 hps = DictToAttrRecursive(sovits_dict["config"])
344 logger.debug(f"SoVITS config: {hps}")

File d:\CondaEnv\envs\normal\lib\site-packages\torch\serialization.py:1026, in load(f, map_location, pickle_module, weights_only, mmap, **pickle_load_args)
1024 except RuntimeError as e:
1025 raise pickle.UnpicklingError(UNSAFE_MESSAGE + str(e)) from None
-> 1026 return _load(opened_zipfile,
1027 map_location,
1028 pickle_module,
1029 overall_storage=overall_storage,
1030 **pickle_load_args)
1031 if mmap:
1032 raise RuntimeError("mmap can only be used with files saved with "
1033 "`torch.save(_use_new_zipfile_serialization=True), "
1034 "please torch.save your checkpoint with this option in order to use mmap.")

File d:\CondaEnv\envs\normal\lib\site-packages\torch\serialization.py:1438, in _load(zip_file, map_location, pickle_module, pickle_file, overall_storage, **pickle_load_args)
1436 unpickler = UnpicklerWrapper(data_file, **pickle_load_args)
1437 unpickler.persistent_load = persistent_load
-> 1438 result = unpickler.load()
1440 torch._utils._validate_loaded_sparse_tensors()
1441 torch._C._log_api_usage_metadata(
1442 "torch.load.metadata", {"serialization_id": zip_file.serialization_id()}
1443 )

File d:\CondaEnv\envs\normal\lib\site-packages\torch\serialization.py:1431, in _load..UnpicklerWrapper.find_class(self, mod_name, name)
1429 pass
1430 mod_name = load_module_mapping.get(mod_name, mod_name)
-> 1431 return super().find_class(mod_name, name)

ModuleNotFoundError: No module named 'utils'

[shibing624]

参考示例代码 https://github.com/shibing624/parrots/blob/master/examples/demo_tts.py ，把

utils的引用路径带上。

import parrots
from parrots import TextToSpeech

parrots_path = parrots.__path__[0]
sys.path.append(parrots_path)  # add parrots to sys.path

原因：原gpt-sovits训练时引入的坑，用的torch.save保存的模型，不仅保存了权重，还有模型结构，所以加载模型时也需要引入。

[bombhero]

多谢大佬回复,已经解决.
另外, https://github.com/shibing624/parrots/blob/master/parrots/tts.py#L201
添加encoding参数.建议修改为
with open(filepath, 'r', encoding='UTF-8') as file:
因为config.json里面有中文,我在vscode里面用jupyter执行不会报gbk的错.但写成py文件,执行时候就报gbk的问题了.
再次感谢大佬的回复

[shibing624]

done